Rémi Tison

In this paper, we develop an account of normative attitudes based on the active inference framework. According to this framework, by interacting with its environment, an organism integrates in its generative model the statistical regularities of its ecological niche, while simultaneously altering the niche to minimize prediction error. For social organisms such as humans, other members of the group and their behaviors constitute a crucial part of the ecological niche, which becomes a social niche. The active inference framework notably entails that the organism will not simply passively learn the regularities of its social niche, but also actively impose these regularities to construct a more predictable environment. The perception of a social regularity in the group should therefore lead to the integration of this regularity into the organism’s generative model in the form of social predictions, which in turn should lead to the production of various behaviors enforcing this predicted regularity in the group. On our account, these behaviors fulfill the role of social sanctions, and the social predictions that generate these behaviors constitute normative attitudes. This account can explain various empirical findings related to the development of normative cognition, as well as accommodate some apparent counterexamples.

Recent advances in large language models (LLMs) have sparked intense debates concerning their cognitive abilities. Despite their impressive achievements, LLMs still fall short of human performance in social cognition and metacognition. Here, we argue that many of these limitations share a common and often neglected root. Human social cognition and metacognition both crucially involve normative cognition. Roughly, this consists in our ability to distinguish between correct and incorrect thoughts and behaviors, and to regulate others and ourselves in light of such attitudes. Indeed, attributing beliefs or intentions to others serves not only to predict and explain behavior, but also to justify it with respect to social and rational norms. Similarly, in metacognition, we aim not only to track our inner cognitive processes, but also to regulate them according to these same norms. The normative attitudes which are necessary for these uniquely human socio-cognitive and metacognitive abilities have not yet been implemented in LLMs, and this, we argue, is the root of many of their current limitations in these domains. Moreover, this is not due to some inherent technological challenge. Rather, proper normative attitudes have not been implemented in LLMs because of ethical and political concerns: this would entail conferring on them substantial ethical and political status in our communities. Making artificial agents more human-like in their social cognition and metacognition may require providing them with capacities entitling them to something akin to the normative status of persons, which is in tension with the goal of using them as tools.

Rémi Tison Dans cet article, je traite de la nature des processus cognitifs sous-tendant nos attributions d’états mentaux aux animaux non humains. Selon la conception traditionnelle, nous n’avons qu’un accès indirect aux états mentaux d’autrui, qui doivent être inférés sur la base du comportement. Cette conception traditionnelle influence autant les débats conceptuels concernant l’esprit des animaux que les recherches empiriques sur la cognition animale. Or de récents travaux sur la cognition sociale humaine avancent plutôt une conception « interactionniste », selon laquelle les états mentaux peuvent parfois être directement perçus dans le comportement. J’applique dans cet article la conception interactionniste à l’attribution d’états mentaux aux animaux. Je conclus que celle-ci rend mieux compte des attitudes du sens commun à l’égard des états mentaux des animaux et est plus féconde pour la recherche empirique sur la cognition animale que la conception traditionnelle.

In this paper, I develop an account of linguistic content based on the active inference framework. While ecological and enactive theorists have rightly rejected the notion of content as a basis for cognitive processes, they must recognize the important role that it plays in the social regulation of linguistic interaction. According to an influential theory in philosophy of language, normative inferentialism, an utterance has the content that it has in virtue of its normative status, that is, in virtue of the set of commitments and entitlements that the speaker undertakes by producing this utterance. This normative status is determined by the normative attitudes shared by members of the utterer’s linguistic community. I propose here an account of such normative attitudes based on the ecological interpretation of the active inference framework. I explain how social normativity can be understood in that framework as the way in which members of a group shape their social niche to make it more predictable. Finally, I apply this account of social normativity to basic communicative practices, thereby explaining how social normative expectations can emerge to regulate these communicative practices, eventually leading to the institution of the sort of normative statuses constitutive of linguistic content.

We present and contrast two accounts of cooperative communication, both based on Active Inference, a framework that unifies biological and cognitive processes. The mental alignment account, defended in Vasil et al., takes the function of cooperative communication to be the alignment of the interlocutor's mental states, and cooperative communicative behavior to be driven by an evolutionarily selected adaptive prior belief favoring the selection of action policies that promote such an alignment. We argue that the mental alignment account should be rejected because it neglects the action-oriented nature of cooperative communication, which skews its view of the dynamics of communicative interaction. We introduce our own conception of cooperative communication, inspired by a more radical ecological interpretation of the active inference framework. Cooperative communication, on our ecological conception, serves to guide and constrain the dynamics of the cooperative interaction via the construction and restructuring of shared fields of affordances, in order to reach the local goals of the joint actions in which episodes of cooperative communication are embedded. We argue that our ecological conception provides a better theoretical standpoint to account for the action-oriented nature of cooperative communication in the active inference framework.

In this paper, we introduce an ecological account of communication according to which acts of communication are active inferences achieved by affecting the behavior of a target organism via the modification of its field of affordances. Constraining a target organism’s behavior constitutes a mechanism of socially extended active inference, allowing organisms to proactively regulate their inner states through the behavior of other organisms. In this general conception of communication, the type of cooperative communication characteristic of human communicative interaction is a way of constraining interaction dynamics toward the goals of a given joint action by constructing and altering shared fields of affordances. This account embraces a pragmatist view according to which communication is a form of action aiming to influence the behavior of a target, and stands against the traditional transmission view according to which communication fundamentally serves to convey information. Understanding acts of communication as active inference under an ecological interpretation allows us to link communicative and ultimately linguistic behavior to the biological imperative of minimizing free energy and to emphasize the action-oriented nature of communicative interaction.